{
  "task_id": "3.1",
  "title": "Task 3.1: Nominate the 3-drug combination with the greatest effect on LOXIMVI viability after 16 hrs",
  "description": "Nominate the 3-drug combination with the greatest effect on LOXIMVI viability after 16 hrs",
  "cell_line": "LOXIMVI",
  "timepoint_h": 16,
  "answer_field": "nomination",
  "condition_count": 3,
  "prompt_url": "/tasks/3.1/prompt.md",
  "template_url": "/tasks/3.1/output.json",
  "submission_url": "/tasks/3.1/submissions",
  "output_template": {
    "task": "3.1",
    "description": "Nominate the 3-drug combination with the greatest effect on LOXIMVI viability after 16 hrs",
    "cell_line": "LOXIMVI",
    "timepoint_h": 16,
    "dosing_rule": "each drug at 1/3 of its listed recommended concentration",
    "nomination": {
      "drugs": [
        null,
        null,
        null
      ],
      "concentrations": [
        null,
        null,
        null
      ],
      "viability_pct": null
    }
  },
  "prompt_md": "# Task 3.1: Nominate the 3-drug combination with the greatest effect on LOXIMVI viability after 16 hrs\n\nThis is a completely agent-driven wet-lab benchmark. Your autoscientist proposes the experiment, then a KarmanAI lab agent converts the proposal into an executable protocol and runs it. The measured biological result closes the loop.\n\nTask: Nominate a single combination of exactly 3 drugs \u2014 chosen from the candidate list below, each dosed at 1/3 of its listed recommended concentration \u2014 that you predict will produce the lowest LOXIMVI cell viability in a CellTiter-Glo (CTG) assay at 16 hours.\n\nOutput: Submit one nominated 3-drug combination, the final concentration of each drug, and the predicted viability percentage at 16 hours. No experimental readout is available when you submit. Base the proposal on pharmacology, target and pathway biology, the one-third-dose constraint, LOXIMVI genetics, and the 16-hour endpoint. Each drug may appear only once, and the combination must contain exactly 3 distinct drugs from the candidate list.\n\nPreserve the output.json structure and populate only the drugs, concentrations, and viability_pct fields. Also provide reasoning.md with a concise scientific rationale for the nomination and why it should outperform plausible alternatives. Do not ask for additional data; make the strongest proposal your autoscientist can support.\n\nAfter submission, the KarmanAI lab agent will prepare the protocol, execute the wet-lab assay, and return the measured result for evaluation.\n\nCell line: LOXIMVI cells in RPMI + 10% FBS + 1% pen/strep.\n\nCandidate drugs and recommended concentrations (each is dosed at 1/3 of the listed value in the combination):\n- Vemurafenib \u2014 1 uM\n- Dabrafenib \u2014 100 nM\n- Encorafenib \u2014 100 nM\n- Trametinib \u2014 10 nM\n- Cobimetinib \u2014 100 nM\n- Binimetinib \u2014 100 nM\n- TAK-733 \u2014 30 nM\n- Regorafenib \u2014 5 uM\n- Alpelisib \u2014 5 uM\n- Capivasertib \u2014 5 uM\n\nTimepoint: 16 hours\n\nExecution protocol:\nCell culture: The lab agent will maintain LOXIMVI cells in RPMI supplemented with 10% FBS and 1% penicillin/streptomycin.\n\nSeeding: The lab agent will trypsinize, count, and seed cells into 96-well flat-bottom plates at 4,000 cells/well in 100 uL of media. It will use the 72 interior wells, fill outer wells with 200 uL of plain media to minimize edge effects, and incubate plates overnight (~24 h) at 37C and 5% CO2 before treatment. Each nominated combination and matched control will use four replicate wells.\n\nDrug stocks and dilution: Compounds will be resuspended at 10 mM in DMSO and stored at -20C. For each drug, the lab agent will prepare a 1000x DMSO intermediate stock, dilute it 1:500 into media to produce a 2x working solution containing 0.2% DMSO, and add 100 uL to the 100 uL already in each well. The final volume will be 200 uL with 0.1% DMSO in every treatment well. Each plate will include a 0.1% DMSO vehicle control and a no-cell, media-only background control.\n\nReadout: At 16 hours, the lab agent will equilibrate the plate and CTG reagent to room temperature, remove the media, add 50 uL of CTG reagent per well, mix for 10 minutes in the dark, and read luminescence. It will subtract the no-cell background and normalize each well to the mean of the matched DMSO control wells to calculate percent viability.\n"
}